"ERMI 2" Status Report, as of May 16th, 2007
The
final report of the DLF’s Electronic Resources Management Initiative
(“ERMI”) was completed in August 2004 and has been instrumental in
spurring development of ERM systems and services – but in the opinions of
its authors, it left some important “loose ends.” To tie up those
loose ends, a second phase to the project – called “ERMI 2”
– has been underway with DLF sponsorship since November 2005 (see
http:www.diglib.org/standards/dlf-erm05.htm). Work on all facets of ERMI 2 is
now nearing completion, and release of a revised set of deliverables and final
report is planned for this summer. What follows is a brief status report on the
most important components of the project.
- E- Resource
Usage Statistics. The most significant
deliverable for this segment of ERMI 2 was to be development of “a
protocol for automated delivery of COUNTER-compliant vendor usage data to
ERM systems, and a demonstration of its use in practice.” Envisioned
as a way to dramatically reduce the time spent by library staff in
gathering and reporting usage data, this idea has been warmly embraced by
the library community, and the protocol itself has been developed and
publicized as SUSHI (for “Standardized Usage Statistics Harvesting
Initiative”). Sponsored and supported by NISO (see http://www.niso.org/committees/SUSHI/SUSHI_comm.html),
and co-chaired by ERMI 2 steering group member Adam Chandler and
EBSCO’s Oliver Pesch. SUSHI is well on its way to becoming an
official standard, since it is now in the “draft standard for trial
use” phase of development, with a comment deadline of May 20th.
SUSHI has been tested successfully and continues to get a lot of
attention, since its potential for reducing library staff time spend
gathering statistics appears now to be well established. For example, at a
recent ICOLC meeting in Montreal, Kathy Perry of the VIVA consortium
described a pilot project in which use of SUSHI reduced the time to
process the reports for one vendor from 10 hours to 15 minutes.
The adoption of SUSHI by content providers will be
critical to its success, and may be accelerated if a recommendation to include
SUSHI as a requirement for COUNTER compliance is accepted for the next release
of the COUNTER Code of Practice -- due out in early 2008. COUNTER is also
working to revamp the consortium reports in the next release as well. These
reports will be available only in XML (directly or through SUSHI) and will
allow a consortium to retrieve detailed usage information for each member in a
single message.
- License
expression. Dozens of the 300+ data
elements included in the ERMI Data Dictionary published as part of the
original project report dealt with “terms of use” from
licenses. However, these data elements and proposed values had yet to be
tested widely by libraries. Accordingly ERMI 2 aimed to foster uptake and
testing of the “ERMI approach” to the description and sharing
of license information (“license mapping”), while working with
NISO, EDItEUR and other appropriate publisher, vendor, and library groups
to take a more rigorous and systematic approach to reviewing the
suitability of the ERMI license element set as a basis for standardized
description and communication of e-resource license provisions and related
licensing metadata.
- Training in
License Term Mapping. A pilot license mapping
workshop was cosponsored by the Association of Research Libraries and the
DLF and conducted in June 2005, and the ERMI 2 plan called for further
development of the pilot course materials and provision of additional
training opportunities. Trisha Davis (Ohio State University) and Diane
Grover (University of Washington) were both named ARL Visiting Program
Officers to undertake this work, and they began their one-year
appointments in January 2006. Since then, 4 well-attended 3-4 hour
workshops have been presented to upwards of 160 people (at NASIG, ALA
Annual and the Charleston Conference in 2006, and ALA midwinter in 2007),
the instructional materials have been refined, and valuable experience
and attendee feedback gained with each offering. That feedback is being
used to inform additional license expression work described below; common
and consistent feedback themes were that interpreting even relatively
simple and “library-friendly” licenses can be quite difficult
and complex, that there is a real need for simplicity in expressing the
important terms of use, and that there is an ongoing need for training
both in licensing skills and mapping.
- License
Expression Data Standards. Prior to and independent of
the development of the ERMI data dictionary, work toward a more detailed
and differently structured licensing data dictionary had been undertaken
under EDItEUR sponsorship, in conjunction with ONIX. Discussions of
possible areas and strategies for collaboration across the two projects
started to take place during the initial phase of ERMI, which helped pave
the way toward formation of the NISO License Expression Working Group,
co-chaired by ERMI steering group member Nathan Robertson and Alicia Wise
from the Publishers Licensing Society. (see http://www.niso.org/committees/License_Expression/LicenseEx_comm.html).
By December a draft mapping of the ERMI license terms to the ONIX set
(now referred to as the ONIX Publications License, or
“ONIX-PL”) was produced with partial support from the DLF,
and discussed in detail at meetings in Boston in late December attended
by representatives from ERMI, NISO, ILS vendors, ONIX, and others.
The mapping is currently in process of being
finalized, but a few important issues have been identified. The metadata
crosswalk will be imperfect. It will be possible to map only a selection
of concepts from ONIX-PL to ERMI. Furthermore, during the ONIX-PL to ERMI
mapping process, certain values will be converted to notes. Any return
mapping from ERMI to ONIX-PL will leave these concepts in note form.
Specificity and granularity is therefore lost in the mapping
process. In addition, the ONIX-PL approach allows for substantially
more detailed license expression than many libraries need or want.
Although ONIX-PL cannot be used to apply DRM technologies, many librarians
continue to be concerned that the creation of detailed rights encodings will
facilitate the application of automated restrictions in the future.
Moreover, there is a continuing question from libraries about the cost in time
and effort of developing detailed expressions of individual licenses when
libraries do not need or want that level of detail. Lastly, while
it might be desirable for libraries and publishers or vendors to exchange
versions of licenses during a negotiation process, ERMs would have to be
substantially re-tooled to support that from the libraries side. And as
these developments have been taking place, NISO’s Shared E-Resource
Understanding initiative ("SERU": http://www.niso.org/committees/SERU/index.html),
which aims to offer an alternative to current reliance on negotiated licenses,
has gathered momentum. While it is too early to tell what effect SERU
might have on license expression and mapping, reducing the number of licenses
to be analyzed and mapped, and promoting simpler agreements would both be
welcomed by many librarians.
- Interoperability. An
earlier ERMI 2 status report noted that “. . . the final area of
focus for the project is interoperability between ERM systems and other
ILS modules, which is problematic for those libraries in
“mixed” environments where an ERM has been acquired from one
vendor and it must be made to work with and complement another
vendor’s ILS – particularly an acquisition module.” Last
summer a small group under the leadership of Norm Medeiros (Haverford
College and the Tri-College Consortium near Philadelphia) began examining
this problem, and produced a draft white paper that was released for
comment in January. The white paper is now undergoing revision, and when
complete in June will include use cases and a list of data elements
perceived as “core” to interoperability, provide comments
supporting the need for standard “resource” identifiers, and
discuss some emerging issues concerning cost per use calculations and
their interpretation.
An
important goal of all ERMI 2 work clearly has been the development and wide
adoption of standards for the project’s areas of interest: delivery of
electronic usage statistics, license expression, and ERMI/Acquisitions module
interoperability. The initiative’s final report will offer suggestions
for further progress toward that goal.
Tim
Jewell, ERMI 2 Project Coordinator
Director, Information Resources, Collections
and Scholarly Communication
University of Washington Libraries, Box 352900
Seattle, WA 98195-2900
phone: 206-543-3890 fax: 206-685-8727
[log in to unmask]